[feature] Add a deploy to cloud run component

[stephenleo]

Feature Area

/area components

What feature would you like to see?

Google Cloud Run is a fantastic way to deploy containers in a serverless manner. Cloud Run recently introduced GPU support which makes it attractive for deploying ML models completely serverless. This feature is to create a new google-cloud-pipeline-component to deploy a trained ML model from Model Registry to Cloud Run.

I have built a custom component that I've been using for my personal projects that deploys a model from Model Registry to Cloud Run. I'd like to polish it up and submit a PR. The component is pretty self-contained and shouldn't impact any other aspect of the framework.

What is the use case or pain point?

1. Cloud Run has all the benefits of serverless: Easy to manage, cheap, scale to zero
2. Feature parity with Sagemaker: Sagemaker provides serverless endpoints while VertexAI doesn't natively support serverless endpoints. Cloud Run bridges the gap.

Is there a workaround currently?

None

---

Love this idea? Give it a 👍.

[google-oss-prow]

@rimolive: The label(s) /label platform/gcp cannot be applied. These labels are supported: tide/merge-method-merge, tide/merge-method-rebase, tide/merge-method-squash, lifecycle/needs-triage. Is this label configured under labels -> additional_labels or labels -> restricted_labels in plugin.yaml?

In response to this:

/label platform/gcp

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

[rimolive]

/cc @kubeflow/google